Related Experiment Video
Updated: Aug 5, 2026
Fine-Tuning Large Language Models Using Entity Hallucination Index for Text Summarization
Published on: January 9, 2026
Proposals for the #Enzian addendum and update: an international expert survey for item generation
Milena Petrovic1, Jörg Keckstein2, Mohamed Mabrouk1
1UCLH Endometriosis Centre, University College London Hospital, London, United Kingdom.
Background:
Accurate and standardised characterisation of endometriosis extent is essential for patient counselling regarding pain and fertility outcomes, surgical planning, multidisciplinary communication, and clinical research. The Enzian classification has been applied in Central Europe for over two decades, while the updated #Enzian classification, published in 2021, has subsequently gained broader international acceptance.
Objectives:
To evaluate international expert perspectives and generate potential items informing updates to selected compartments of the #Enzian classification.
Methods:
Twenty-six internationally recognised experts in endometriosis surgery, imaging, and reproductive medicine were invited to participate in a structured survey. The electronic questionnaire comprised 29 single-choice and open-ended items addressing five #Enzian compartments: T (tuboovarian condition), B (sacrouterine ligaments, cardinal ligaments, pelvic sidewall), C (rectum), FB (bladder), and FA (adenomyosis). Responses were collected anonymously, analysed descriptively, and thematically grouped. Consensus thresholds followed established criteria.
Main Outcome Measures:
Expert perspectives regarding the need for revision of selected #Enzian compartments and proposed modifications.
Results:
Twenty-three experts responded (88.5%). Strong agreement for revision was observed for compartments B, C, and FB (>80%). Moderate agreement (>60%) supported updates to compartments T and FA. Specific suggested refinements included adding the suffix d within compartment T to reflect tubal dilatation (73.9% in favour) and the suffix P within compartment B to denote pelvic sidewall/parametrial involvement (78.3% in favour). For colorectal disease (compartment C), respondents supported reporting lesion-anal verge distance and multifocality. A graded FA1-FA3 system for adenomyosis was favoured (73.9%). Opinions regarding descriptors of bladder endometriosis were heterogeneous.
Conclusions:
Experts identified clear priorities for refining the #Enzian classification, particularly regarding parametrial involvement, colorectal deep endometriosis and adenomyosis extent. These findings could provide a structured framework to inform subsequent Delphi consensus rounds.
What Is New?:
The study identifies international expert-agreed priorities for refinement of selected #Enzian classification compartments.